For centuries humankind has been brewing craft beers, which are typically associated with craftmanship and specific skills and knowledge to create the perfect beer. Beer breweries continuously introduce an unending list of different kinds of beers, leaving the consumer to wonder how to make sense out of all these options.
Instead of continuously introducing new beers onto the market, it is now a company’s goal to improve the craft beers that already exist. IntelligentX is the first beer brewery that uses Automated Brewing Intelligence (ABI) to create craft beers that are continuously improved based on consumer feedback.
Programmers transcribe brewer’s intuition and knowledge into code that Artificial Intelligence can understand. Different ingredients and flavors are also added to the Artificial Intelligence system, which allows for all different kinds of craft beers to be created. Moreover, a special code is printed on the bottles that redirects consumers that scan the code to a chatbot which asks for feedback. The chatbot learns which questions to ask and adapts feedback questions accordingly in order to improve the beers. This way the system can learn the taste preferences from different kind of consumers and make appropriate adjustments. ABI has been able to create sixteen different craft beers from four different kind of beer styles.
In my opinion, IntelligentX is a perfect example on how everyday products can be improved by the use of new technologies such as AI. Furthermore, once these kinds of systems can be introduced throughout different industries, this would eliminate many new launched product launches. By integrating the end-consumer into the creation process consumers are more connected to the brand and the products they carry. I believe there is an increasing need for brand and people to talk, AI systems such as ABI can take on this challenge.
The implementation of AI can go much further than just improving beer, therefore, my question to you is in which products or industries do you think similar improvement can be made possible by the introduction of AI? Would you order an IntelligentX beer the next time you are in a bar, why yes or no?
Hi Iris,
Thank you for the insightful and engagin post!
Before I go into your questions, I have a questin for you: do you know if IntelligentX is profitable?
And do you know if they purely depend on AI insight, or if they supplement their model with insights or hints?
I do think that AI can be a fruitful measure for creating all sorts of new – taste based – products.
There have been experiments in the creation of classical music with AI that were highly succesful. In fact, music critics were invited to listen to an AI created part and a human created part and they were unable to distinct the two.
However, I do not think AI will ever be able to completely replace human ingenuity.
AI is commonly trained on large data sets. These data sets depend on human created content.
I am very skeptical about the question whether AI can ever create “something” out of “nothing”
Paul
Hi Paul,
Thank you for your comment and sharing your view on IntelligentX. The company was founded in 2016 and has spend the last year on improving the beers and collecting user feedback. Therefore, it is not yet profitable. As far as I could find information the ABI truly learns from feedback and machine learning and no human interference is used.
I do agree with you on your view of humans not being completely replaced. However, I feel that once these systems become so smart they easily outperform humans and we might only be there to feed the machines data rather than modifying outcomes.
I think it is very cool that AI is now also used in creating music, which I didn’t know. Thanks for sharing that example Paul.
Beer is already perfect, you can’t improve it! Haha, just joking, but I really think that this can be a very hard task, as, without counting a small percentage of the population that really are beer connoisseurs, I think most of us don’t really know enough about beer to be able to give useful feedback – this means that the questions asked must be very good to actually receieve good information. The best part of this project is probably exactly the fact that the chatbot learns what questions to ask – so maybe it can actually work!
Also I imagine that at some point the system will have enough knowledge about your preferences to start offering you beers that you might like, which might be another interesting use for it.
I would definitely order an IntelligentX beer the next time I am in a bar – at least to try it out. Have to decide after if I will continue using it 😉
Hi Stoyko,
Thanks for taking the time to leave a comment.
I agree with your comment, not everyone is a beer expert, therefore, their opinions and feedback should not always be integrated. The start-up has been working or a year on training their chatbot & ABI. In the first few weeks for example the chatbot would ask a consumer whether they wanted to more alcohol in their beer, to which most of them said yes. Once the beer was improved according to the feedback most consumers actually didn’t like the higher percentage of alcohol. With this feedback the chatbot is able to improve the questions that are asked, instead of asking for alcohol percentages, they could ask if the consumer likes a stronger beer taste.
After receiving a lot of feedback, the ABI & chatbot know who’s opinion to value.
I also believe that implementing beer advice would be an added value!